Clustering Spotify Tracks
Alfan
3/8/2020
Intro
Background
What is spotify? Spotify is a digital music, podcast, and video streaming service that gives you access to millions of songs and other content from artists all over the world. Recently spotify is one of biggest digital music and podcast service in world.
Spotify definitely is one of tech company has very advanced technology. One of example is, each of song track have uploaded to platform, they will identified. We can get audio feature information for each track, and access very easy we can use this Api link. In this case we will using spotify dataset from API from this source kaggle.
We will try to analyze popularity for each track we get, based on data we will try to find there is relation from popularity with other feature or variable. We will also try to do clustering analysis using K-means method and for sure we will find try to reduction dimensionality using Principle Component Analysis (PCA)
Dataset
We will use dataset we get from kaggle, can download from this source source)
Intial Setup and Library
# Starting collection for data science
library(tidyverse)
# Processing string
library(glue)
# Processing date data type
library(lubridate)
# Multivariate Data Analyses
library(factoextra)
# Multivariate Data Analyses
library(FactoMineR)
# Data visualization
library(ggplot2)
library(viridis)
library(GGally)
library(scales)Import Data
The dataset we download from kaggle, we will import dataset. This dataset cotontaint audio feature from a track.
tracks <- read_csv("data/SpotifyFeatures.csv")Observe structure and preview imported dataset
glimpse(tracks)## Observations: 232,725
## Variables: 18
## $ genre <chr> "Movie", "Movie", "Movie", "Movie", "Movie", "Movie"…
## $ artist_name <chr> "Henri Salvador", "Martin & les fées", "Joseph Willi…
## $ track_name <chr> "C'est beau de faire un Show", "Perdu d'avance (par …
## $ track_id <chr> "0BRjO6ga9RKCKjfDqeFgWV", "0BjC1NfoEOOusryehmNudP", …
## $ popularity <dbl> 0, 1, 3, 0, 4, 0, 2, 15, 0, 10, 0, 2, 4, 3, 0, 0, 0,…
## $ acousticness <dbl> 0.61100, 0.24600, 0.95200, 0.70300, 0.95000, 0.74900…
## $ danceability <dbl> 0.389, 0.590, 0.663, 0.240, 0.331, 0.578, 0.703, 0.4…
## $ duration_ms <dbl> 99373, 137373, 170267, 152427, 82625, 160627, 212293…
## $ energy <dbl> 0.9100, 0.7370, 0.1310, 0.3260, 0.2250, 0.0948, 0.27…
## $ instrumentalness <dbl> 0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.12…
## $ key <chr> "C#", "F#", "C", "C#", "F", "C#", "C#", "F#", "C", "…
## $ liveness <dbl> 0.3460, 0.1510, 0.1030, 0.0985, 0.2020, 0.1070, 0.10…
## $ loudness <dbl> -1.828, -5.559, -13.879, -12.178, -21.150, -14.970, …
## $ mode <chr> "Major", "Minor", "Minor", "Major", "Major", "Major"…
## $ speechiness <dbl> 0.0525, 0.0868, 0.0362, 0.0395, 0.0456, 0.1430, 0.95…
## $ tempo <dbl> 166.969, 174.003, 99.488, 171.758, 140.576, 87.479, …
## $ time_signature <chr> "4/4", "4/4", "5/4", "4/4", "4/4", "4/4", "4/4", "4/…
## $ valence <dbl> 0.8140, 0.8160, 0.3680, 0.2270, 0.3900, 0.3580, 0.53…Variable Explaination:
1. genre : Track genre
2. artist_name : Artist name
3. track_name : Title of track
4. track_id : The Spotify ID for the track.
5. popularity : Popularity rate (1-100)
6. acousticness : A confidence measure from 0.0 to 1.0 of whether the track is acoustic. 1.0 represents high confidence the track is acoustic.
7. danceability : Danceability describes how suitable a track is for dancing based on a combination of musical elements including tempo, rhythm stability, beat strength, and overall regularity. A value of 0.0 is least danceable and 1.0.
8. duration_ms : The duration of the track in milliseconds.
9. energy : Energy is a measure from 0.0 to 1.0 and represents a perceptual measure of intensity and activity. Typically, energetic tracks feel fast, loud, and noisy. For example, death metal has high energy, while a Bach prelude scores low on the scale.
10. instrumentalness : Predicts whether a track contains no vocals. “Ooh” and “aah” sounds are treated as instrumental in this context. Rap or spoken word tracks are clearly “vocal”. The closer the instrumentalness value is to 1.0, the greater likelihood the track contains no vocal content. Values above 0.5 are intended to represent instrumental tracks, but confidence is higher as the value approaches 1.0.
11. key : The estimated overall key of the track. Integers map to pitches using standard Pitch Class notation . E.g. 0 = C, 1 = C♯/D♭, 2 = D, and so on. If no key was detected, the value is -1.
12. liveness : Detects the presence of an audience in the recording. Higher liveness values represent an increased probability that the track was performed live. A value above 0.8 provides strong likelihood that the track is live.
13. loudness : The overall loudness of a track in decibels (dB). Loudness values are averaged across the entire track and are useful for comparing relative loudness of tracks.
14. mode : Mode indicates the modality (major or minor) of a track, the type of scale from which its melodic content is derived. Major is represented by 1 and minor is 0.
15. speechiness : Speechiness detects the presence of spoken words in a track. The more exclusively speech-like the recording (e.g. talk show, audio book, poetry), the closer to 1.0 the attribute value. Values above 0.66 describe tracks that are probably made entirely of spoken words. Values between 0.33 and 0.66 describe tracks that may contain both music and speech, either in sections or layered, including such cases as rap music. Values below 0.33 most likely represent music and other non-speech-like tracks.
16, tempo : The overall estimated tempo of a track in beats per minute (BPM). In musical terminology, tempo is the speed or pace of a given piece and derives directly from the average beat duration.
17. time_signature : An estimated overall time signature of a track. The time signature (meter) is a notational convention to specify how many beats are in each bar (or measure).
18. valence : A measure from 0.0 to 1.0 describing the musical positiveness conveyed by a track. Tracks with high valence sound more positive (e.g. happy, cheerful, euphoric), while tracks with low valence sound more negative (e.g. sad, depressed, angry).
Data Wrangling
First of all, we would check NA or Empty value of each variable, We didnt find any NA inside data
colSums(is.na(tracks))## genre artist_name track_name track_id
## 0 0 0 0
## popularity acousticness danceability duration_ms
## 0 0 0 0
## energy instrumentalness key liveness
## 0 0 0 0
## loudness mode speechiness tempo
## 0 0 0 0
## time_signature valence
## 0 0Some variable have wront type data, we need to convery the data type:
genre: to factor
key: to factor
genre: to factor
mode: to factor
tracks <- tracks %>%
mutate(genre = as.factor(genre),
key = as.factor(key),
genre = as.factor(str_replace_all(genre, "[[:punct:]]", "")),
mode = as.factor(mode))Drop variable that we think didnt related with our case. In this case we prioritize variable with numerical data type.based on summary we will drop track id, time_signature, track_name.
summary(tracks)## genre artist_name track_name
## Childrens Music: 14756 Length:232725 Length:232725
## Comedy : 9681 Class :character Class :character
## Soundtrack : 9646 Mode :character Mode :character
## Indie : 9543
## Jazz : 9441
## Pop : 9386
## (Other) :170272
## track_id popularity acousticness danceability
## Length:232725 Min. : 0.00 Min. :0.0000 Min. :0.0569
## Class :character 1st Qu.: 29.00 1st Qu.:0.0376 1st Qu.:0.4350
## Mode :character Median : 43.00 Median :0.2320 Median :0.5710
## Mean : 41.13 Mean :0.3686 Mean :0.5544
## 3rd Qu.: 55.00 3rd Qu.:0.7220 3rd Qu.:0.6920
## Max. :100.00 Max. :0.9960 Max. :0.9890
##
## duration_ms energy instrumentalness key
## Min. : 15387 Min. :0.0000203 Min. :0.0000000 C :27583
## 1st Qu.: 182857 1st Qu.:0.3850000 1st Qu.:0.0000000 G :26390
## Median : 220427 Median :0.6050000 Median :0.0000443 D :24077
## Mean : 235122 Mean :0.5709577 Mean :0.1483012 C# :23201
## 3rd Qu.: 265768 3rd Qu.:0.7870000 3rd Qu.:0.0358000 A :22671
## Max. :5552917 Max. :0.9990000 Max. :0.9990000 F :20279
## (Other):88524
## liveness loudness mode speechiness
## Min. :0.00967 Min. :-52.457 Major:151744 Min. :0.0222
## 1st Qu.:0.09740 1st Qu.:-11.771 Minor: 80981 1st Qu.:0.0367
## Median :0.12800 Median : -7.762 Median :0.0501
## Mean :0.21501 Mean : -9.570 Mean :0.1208
## 3rd Qu.:0.26400 3rd Qu.: -5.501 3rd Qu.:0.1050
## Max. :1.00000 Max. : 3.744 Max. :0.9670
##
## tempo time_signature valence
## Min. : 30.38 Length:232725 Min. :0.0000
## 1st Qu.: 92.96 Class :character 1st Qu.:0.2370
## Median :115.78 Mode :character Median :0.4440
## Mean :117.67 Mean :0.4549
## 3rd Qu.:139.05 3rd Qu.:0.6600
## Max. :242.90 Max. :1.0000
## tracks <- tracks %>% select(-c(track_id,time_signature,track_name))Exploratory Data Analysis
From dataset we get, we found we have genre whish is we can group our data base on it. To make us focus on popularity variable, we would select 5 highest average genre. It can inteprate the genre have big distribution from low to highest popularity. We will visualize data:
genre_popularity <- tracks %>% select(popularity, genre) %>% group_by(genre) %>% summarise("average_popularity" = round(mean(popularity)))
ggplot(data=genre_popularity, mapping = aes(x = reorder(genre,average_popularity), y = average_popularity, fill = genre)) +
geom_col() +
coord_flip() +
theme_minimal() +
theme(
legend.position = "none",
) +
labs(
y = "Average popularity",
x = "Genre"
)
Top 5 Genre with highest average popularity is Pop, Rap, Rock, HipHop and Dance. We filter our dataset and select only this 5 genres.
# Filter
tracks <- tracks %>% filter(genre == "Pop" | genre == "Rap" | genre == "Rock" | genre == "HipHop" | genre == "Dance")
# Total row
NROW(tracks)## [1] 45886Clustering Opportunity
Before we use k-mean as method to clustering, we can use simple way to clustering some factor variable with popularity variable. Here we try to viasualize boxplot popularity with key and genre
tracks %>%
ggplot(aes(x = mode, y = popularity, fill = mode)) +
geom_boxplot() +
scale_fill_viridis(discrete = TRUE, alpha=0.6) +
theme_minimal()
tracks %>%
ggplot(aes(x = genre, y = popularity, fill = genre)) +
geom_boxplot() +
scale_fill_viridis(discrete = TRUE, alpha=0.6) +
theme_minimal()
From both bar, we can see in general genre and key didnt have significant relation to popularity. Even so, there is difference in plot popluaritu and genre when track overall using Key A# popularity slighlty higher and more have stable popularity than others.
Other things, we found that Pop genre have more stable popluarity than other 4 genre. So we can consider an opinion, if producer want get more popularity in the spotify platform, we can make tracks song with Pop genre and overall key in tracks using A# key.
We only visualize genre and key before, and we found each correlation with popularity even though, cant significantly which variable can significantly increase popularity. So we will try visualize other numerical variable to see correlation between them.
ggcorr(tracks, low = "blue", high = "red")## Warning in ggcorr(tracks, low = "blue", high = "red"): data in column(s)
## 'genre', 'artist_name', 'key', 'mode' are not numeric and were ignored
It show popularity dont have strong correlation with others any numberical variable. But we found some variable have strong each other, it indicates that this dataset has multicollinearity and might not suitable for various classification algorithms.
To find more interesting and undiscovered pattern in the data, we will use clustering method using the K-means. We will use Principal Component Analysis (PCA) can be performed for this data to produce non-multicollinearity data, while also reducing the dimension of the data and retaining as much as information possible. The result of this analysis can be utilized further for classification purpose with lower computation.
Data Pre-processing
Since we will implement K-means method and using PCA, its we will perform pre-processing data
hist(tracks$popularity)
set.seed(100)
tracks_sample <- sample_n(tracks, (nrow(tracks) * 0.05))
NROW(tracks_sample)## [1] 2294hist(tracks_sample$popularity)
str(tracks_sample)## Classes 'spec_tbl_df', 'tbl_df', 'tbl' and 'data.frame': 2294 obs. of 15 variables:
## $ genre : Factor w/ 26 levels "A Capella","Alternative",..: 12 9 18 17 12 18 22 17 22 12 ...
## $ artist_name : chr "Talib Kweli" "Andy Grammer" "Mac Miller" "Simple Plan" ...
## $ popularity : num 48 55 50 65 67 82 67 62 52 66 ...
## $ acousticness : num 0.263 0.0409 0.109 0.000491 0.039 0.0776 0.00118 0.154 0.00487 0.73 ...
## $ danceability : num 0.629 0.621 0.438 0.522 0.673 0.643 0.528 0.699 0.542 0.425 ...
## $ duration_ms : num 227373 199113 197852 232067 229507 ...
## $ energy : num 0.787 0.827 0.792 0.751 0.758 0.904 0.858 0.668 0.714 0.406 ...
## $ instrumentalness: num 0 0 0 0.00000222 0 0 0.00000156 0.0000032 0.298 0.00000359 ...
## $ key : Factor w/ 12 levels "A","A#","B","C",..: 11 3 9 5 11 4 11 9 6 8 ...
## $ liveness : num 0.357 0.0815 0.241 0.158 0.341 0.189 0.282 0.362 0.334 0.107 ...
## $ loudness : num -5.62 -7.31 -6.88 -5.46 -3.63 ...
## $ mode : Factor w/ 2 levels "Major","Minor": 1 1 2 1 1 1 1 2 1 2 ...
## $ speechiness : num 0.376 0.0454 0.346 0.0435 0.158 0.0739 0.0493 0.0336 0.028 0.176 ...
## $ tempo : num 85.7 100 175.9 139.5 136 ...
## $ valence : num 0.713 0.65 0.463 0.605 0.542 0.481 0.219 0.314 0.76 0.124 ...tracks_num <- tracks_sample %>% select(-c(genre,artist_name,key,mode))
str(tracks_num)## Classes 'spec_tbl_df', 'tbl_df', 'tbl' and 'data.frame': 2294 obs. of 11 variables:
## $ popularity : num 48 55 50 65 67 82 67 62 52 66 ...
## $ acousticness : num 0.263 0.0409 0.109 0.000491 0.039 0.0776 0.00118 0.154 0.00487 0.73 ...
## $ danceability : num 0.629 0.621 0.438 0.522 0.673 0.643 0.528 0.699 0.542 0.425 ...
## $ duration_ms : num 227373 199113 197852 232067 229507 ...
## $ energy : num 0.787 0.827 0.792 0.751 0.758 0.904 0.858 0.668 0.714 0.406 ...
## $ instrumentalness: num 0 0 0 0.00000222 0 0 0.00000156 0.0000032 0.298 0.00000359 ...
## $ liveness : num 0.357 0.0815 0.241 0.158 0.341 0.189 0.282 0.362 0.334 0.107 ...
## $ loudness : num -5.62 -7.31 -6.88 -5.46 -3.63 ...
## $ speechiness : num 0.376 0.0454 0.346 0.0435 0.158 0.0739 0.0493 0.0336 0.028 0.176 ...
## $ tempo : num 85.7 100 175.9 139.5 136 ...
## $ valence : num 0.713 0.65 0.463 0.605 0.542 0.481 0.219 0.314 0.76 0.124 ...tracks_scale <- scale(tracks_num)fviz_nbclust(tracks_num, kmeans, method = "wss", k.max = 15) +
scale_y_continuous(labels = number_format(scale = 10^(-9), big.mark = ",", suffix = " bil.")) +
labs(subtitle = "Elbow method")
fviz_nbclust(tracks_num, kmeans, method = "silhouette", k.max = 15) 
fviz_nbclust(tracks_num, kmeans, "gap_stat", k.max = 10) + labs(subtitle = "Gap Statistic method")
set.seed(123)
km_tracks <-kmeans(tracks_scale, centers = 2)
km_tracks## K-means clustering with 2 clusters of sizes 736, 1558
##
## Cluster means:
## popularity acousticness danceability duration_ms energy
## 1 0.013047091 0.7995148 -0.13930852 0.12854175 -1.0607826
## 2 -0.006163453 -0.3776912 0.06580942 -0.06072319 0.5011143
## instrumentalness liveness loudness speechiness tempo valence
## 1 0.19655118 -0.2476437 -0.9122186 -0.07286271 -0.2282781 -0.5674398
## 2 -0.09285088 0.1169870 0.4309325 0.03442038 0.1078387 0.2680588
##
## Clustering vector:
## [1] 2 2 2 2 2 2 2 2 2 1 1 2 1 1 2 2 2 2 2 2 2 1 2 1 1 1 2 2 1 1 1 2 2 2 1 2 1
## [38] 1 1 2 2 1 1 2 2 2 2 1 1 1 1 1 1 2 2 2 2 2 2 1 1 1 2 2 2 2 2 1 2 1 2 2 2 2
## [75] 2 1 2 2 2 2 1 1 2 2 2 2 2 1 1 2 2 1 2 2 2 1 2 1 2 1 1 1 2 1 1 2 1 2 1 1 1
## [112] 2 1 2 2 2 1 1 2 1 2 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 1 1 1 2 1 2 2 1 1 2 1 2
## [149] 1 2 2 2 1 1 2 1 1 2 2 1 2 1 2 1 2 2 2 1 1 1 2 2 2 1 2 2 2 2 1 2 1 2 2 2 2
## [186] 2 2 2 2 2 2 2 1 1 2 2 2 1 2 2 2 2 1 2 1 2 2 2 2 2 1 2 1 2 1 2 2 2 2 2 1 2
## [223] 1 2 1 1 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 1 1 2 1 1 2 2 2 2 1 2 1 1 1 2 1 2 1
## [260] 2 1 2 2 2 1 2 1 2 2 2 2 1 2 2 1 1 1 2 1 2 2 2 2 2 1 1 1 2 1 1 1 1 2 2 2 2
## [297] 1 1 2 2 1 1 2 2 2 2 1 1 2 2 1 1 2 2 2 2 2 2 2 1 2 2 2 2 1 2 1 2 1 2 2 2 2
## [334] 1 2 2 2 2 1 2 2 2 2 2 2 2 1 2 1 2 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 1 2 2 1
## [371] 1 1 2 2 2 2 1 1 2 1 2 2 1 2 2 2 2 2 1 1 1 2 2 2 1 2 2 1 2 2 2 2 2 2 1 1 1
## [408] 2 2 1 1 2 2 1 2 2 2 2 1 2 2 1 2 1 2 1 2 1 2 1 2 1 2 2 2 2 2 2 2 2 2 2 1 1
## [445] 1 2 1 2 2 1 2 1 2 1 2 2 2 2 1 1 2 2 2 1 2 1 1 1 2 2 2 1 2 1 1 1 1 2 1 2 2
## [482] 2 1 1 2 2 2 2 1 1 2 1 2 2 2 1 2 2 2 2 2 1 1 2 2 2 2 2 1 2 2 2 1 1 1 2 2 1
## [519] 2 2 1 2 1 2 2 2 2 2 2 1 2 1 1 1 2 1 1 2 1 2 2 2 2 1 2 2 2 1 2 1 2 1 1 2 1
## [556] 2 1 2 1 2 2 1 2 1 2 1 2 1 2 2 1 2 1 2 1 1 2 2 2 2 2 1 1 2 1 2 2 2 1 2 1 1
## [593] 1 2 1 2 2 2 1 1 2 2 2 1 2 1 2 2 2 2 1 1 2 2 2 2 2 2 2 2 2 1 2 1 2 1 2 1 1
## [630] 2 1 2 2 2 1 2 1 2 2 2 2 1 1 1 2 2 1 2 2 2 2 1 2 2 2 2 2 2 2 1 2 2 1 2 2 2
## [667] 2 2 2 2 1 2 1 2 1 2 1 1 1 2 2 1 2 2 2 2 1 2 2 2 1 2 2 2 2 2 2 2 1 2 2 2 1
## [704] 2 2 2 2 2 2 1 2 2 2 2 2 1 2 2 2 2 2 2 1 1 1 1 2 1 2 2 1 2 1 2 2 2 2 2 2 2
## [741] 2 1 2 1 2 1 2 2 2 2 1 2 1 2 2 2 1 2 1 2 1 2 2 2 1 2 2 1 2 1 1 2 2 2 2 2 2
## [778] 1 2 1 1 1 1 2 2 2 2 2 2 1 2 2 2 2 2 1 2 2 2 2 2 2 1 2 2 2 2 1 2 1 2 2 2 1
## [815] 1 2 1 1 2 1 2 2 2 1 2 2 1 1 1 2 2 1 2 2 1 2 2 2 2 1 2 2 2 1 2 2 1 2 1 2 2
## [852] 2 2 1 2 2 1 1 1 2 1 2 2 2 2 2 2 2 2 2 1 2 2 1 1 2 2 2 2 2 2 2 2 1 2 2 1 1
## [889] 2 1 2 2 2 2 1 2 2 2 2 2 1 2 2 1 2 1 1 1 1 2 2 2 1 2 2 2 2 2 2 2 2 1 1 2 1
## [926] 1 1 2 1 2 1 2 2 1 2 2 2 2 1 2 2 1 2 1 2 2 2 2 1 2 2 2 2 2 2 1 2 2 1 2 2 1
## [963] 2 2 2 1 2 2 2 2 1 1 2 2 2 1 2 1 2 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 2 2
## [1000] 2 2 2 2 2 2 1 2 1 2 2 2 2 1 2 2 2 2 2 1 1 1 2 2 2 2 2 2 1 1 2 2 2 2 1 1 2
## [1037] 2 1 2 1 1 1 1 2 2 1 2 1 2 1 2 2 2 2 2 2 1 1 1 2 2 1 2 2 2 1 2 1 2 2 1 1 1
## [1074] 1 2 1 2 1 1 1 2 2 1 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 2 2 1 1
## [1111] 2 2 1 2 1 1 2 2 2 2 1 2 1 2 2 1 1 1 1 2 2 2 1 1 1 1 1 2 1 2 1 1 2 2 2 2 2
## [1148] 2 1 1 1 1 2 2 1 2 2 2 2 1 1 2 2 2 2 2 2 1 2 2 2 1 1 1 1 2 1 2 2 2 1 1 2 2
## [1185] 1 2 1 1 2 2 2 2 1 2 2 2 1 2 1 2 2 2 2 1 2 1 2 2 2 2 2 2 1 2 2 1 2 2 2 2 2
## [1222] 1 2 1 1 1 2 2 1 1 1 2 1 2 2 2 1 2 1 2 2 2 2 2 2 2 2 1 2 1 2 2 1 1 1 2 2 1
## [1259] 1 2 1 2 2 2 2 2 1 1 2 2 2 2 1 2 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 2
## [1296] 2 2 1 2 2 1 2 2 2 2 2 2 2 1 2 1 2 2 2 1 2 2 2 2 1 2 2 2 2 2 2 1 2 1 1 2 2
## [1333] 2 1 2 2 2 1 1 1 2 2 2 1 2 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 1 2 2 2 2 2 2 2 1
## [1370] 2 2 2 2 2 2 1 2 2 2 1 1 2 2 1 2 1 1 2 2 2 2 1 2 1 2 2 2 1 2 1 2 1 2 2 2 2
## [1407] 2 2 2 2 2 2 2 2 1 1 2 2 2 1 1 1 2 2 2 2 2 1 1 1 2 2 2 1 2 2 2 2 2 1 2 1 2
## [1444] 2 2 2 1 2 2 2 2 2 1 2 1 2 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
## [1481] 1 2 2 2 2 2 2 2 1 2 2 1 2 2 1 1 2 1 2 2 2 2 1 2 2 2 1 2 2 2 2 2 2 1 2 2 1
## [1518] 1 1 2 2 1 2 2 1 2 1 1 1 2 1 2 2 2 2 2 2 1 2 1 2 2 2 2 2 2 2 2 2 2 2 1 2 1
## [1555] 2 1 2 2 1 1 1 2 2 2 1 2 2 2 2 1 2 2 1 2 2 2 2 2 2 1 1 1 1 2 2 1 2 1 2 2 2
## [1592] 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 1 1 2 2 2 2 2 2 1 1 2 1 2 2 2 1 1 2 2 2 1
## [1629] 2 1 1 2 2 2 2 1 1 2 2 1 1 1 2 1 2 2 2 1 1 2 1 2 2 2 1 2 1 2 1 2 2 2 2 2 2
## [1666] 1 2 1 1 2 2 2 1 2 1 1 2 1 2 2 1 2 2 1 1 2 1 1 1 2 1 2 2 2 2 2 2 2 1 2 1 2
## [1703] 2 2 2 2 2 2 2 2 2 2 1 1 1 2 1 2 2 2 1 2 1 2 2 2 2 2 2 2 1 2 2 1 1 2 2 2 1
## [1740] 1 2 1 2 2 1 2 2 1 1 2 1 2 1 1 2 1 2 1 2 1 1 1 2 2 2 2 2 1 2 2 2 2 2 2 2 2
## [1777] 2 2 2 1 2 1 1 1 2 2 1 2 1 2 2 1 2 1 2 2 2 2 2 2 2 1 2 1 2 1 2 1 1 1 2 2 1
## [1814] 2 2 2 2 2 2 2 1 2 2 1 1 2 1 1 1 2 2 2 2 2 1 1 1 2 1 2 2 2 2 2 2 1 2 2 2 2
## [1851] 1 2 1 1 2 1 1 1 1 1 1 1 2 1 1 1 2 2 2 1 2 2 2 2 2 2 2 2 1 2 2 1 2 2 2 1 1
## [1888] 2 1 2 2 2 2 1 2 2 2 1 1 2 2 1 1 2 2 2 2 2 2 2 1 2 2 1 2 2 2 2 2 2 2 2 2 2
## [1925] 2 2 2 1 2 2 2 1 2 1 2 2 2 2 2 2 1 1 2 1 2 2 2 2 2 2 2 2 2 2 2 1 1 2 2 2 2
## [1962] 2 1 1 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 1 2 2 2 2 1 2 2
## [1999] 2 2 1 1 2 2 2 2 2 2 2 2 2 2 1 2 1 2 2 2 2 2 2 2 2 2 1 2 2 1 2 1 1 2 2 2 2
## [2036] 1 2 1 2 2 2 2 2 2 1 1 1 2 2 2 2 2 2 1 2 1 2 1 1 2 2 2 2 1 2 1 2 2 2 1 2 2
## [2073] 1 1 2 2 2 1 2 2 2 1 2 2 2 2 2 1 1 2 1 2 1 2 1 1 2 1 2 2 2 2 1 2 1 1 1 2 2
## [2110] 1 2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 1 2 2 2 2 2 2 1 2 1 1 2 2 1 1
## [2147] 2 2 2 1 2 2 2 2 1 2 2 2 2 2 2 1 2 2 1 1 2 2 1 2 1 1 2 2 2 1 2 2 2 1 2 2 1
## [2184] 2 2 1 1 1 1 2 2 2 2 2 1 2 2 2 2 2 2 2 1 1 1 2 2 2 2 2 2 1 2 2 2 2 1 2 2 2
## [2221] 2 2 2 2 2 2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2 2 1 2 2 2 2 2 2 2 1 2 2 1 1
## [2258] 2 2 2 2 2 2 1 2 2 1 2 2 2 1 2 2 1 2 1 1 1 1 2 2 2 2 2 2 2 1 2 2 2 1 2 2 1
##
## Within cluster sum of squares by cluster:
## [1] 8839.964 13010.500
## (between_SS / total_SS = 13.4 %)
##
## Available components:
##
## [1] "cluster" "centers" "totss" "withinss" "tot.withinss"
## [6] "betweenss" "size" "iter" "ifault"tracks_sample$cluster <- as.factor(km_tracks$cluster)
tracks_sample## # A tibble: 2,294 x 16
## genre artist_name popularity acousticness danceability duration_ms energy
## <fct> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 HipH… Talib Kweli 48 0.263 0.629 227373 0.787
## 2 Dance Andy Gramm… 55 0.0409 0.621 199113 0.827
## 3 Rap Mac Miller 50 0.109 0.438 197852 0.792
## 4 Pop Simple Plan 65 0.000491 0.522 232067 0.751
## 5 HipH… Pitbull 67 0.039 0.673 229507 0.758
## 6 Rap Jason Deru… 82 0.0776 0.643 195419 0.904
## 7 Rock Sundara Ka… 67 0.00118 0.528 226395 0.858
## 8 Pop Bazzi 62 0.154 0.699 148230 0.668
## 9 Rock Soda Stereo 52 0.00487 0.542 212893 0.714
## 10 HipH… 6LACK 66 0.73 0.425 286957 0.406
## # … with 2,284 more rows, and 9 more variables: instrumentalness <dbl>,
## # key <fct>, liveness <dbl>, loudness <dbl>, mode <fct>, speechiness <dbl>,
## # tempo <dbl>, valence <dbl>, cluster <fct>tracks_sample %>% ggplot(aes(x = duration_ms, y = popularity, color = cluster)) +
geom_point() +
theme_minimal()
fviz_cluster(object = km_tracks, data = tracks_scale) +
theme_minimal()
non_numeric <- which(sapply(tracks_sample, negate(is.numeric)))
tracks_pca <- PCA(tracks_sample,
scale.unit = T,
quali.sup = non_numeric,
graph = F,
ncp = 15)
summary(tracks_pca)##
## Call:
## PCA(X = tracks_sample, scale.unit = T, ncp = 15, quali.sup = non_numeric,
## graph = F)
##
##
## Eigenvalues
## Dim.1 Dim.2 Dim.3 Dim.4 Dim.5 Dim.6 Dim.7
## Variance 2.309 1.505 1.142 1.027 0.985 0.948 0.847
## % of var. 20.987 13.682 10.379 9.332 8.955 8.619 7.696
## Cumulative % of var. 20.987 34.668 45.047 54.379 63.334 71.954 79.650
## Dim.8 Dim.9 Dim.10 Dim.11
## Variance 0.833 0.645 0.543 0.217
## % of var. 7.575 5.867 4.937 1.971
## Cumulative % of var. 87.225 93.092 98.029 100.000
##
## Individuals (the 10 first)
## Dist Dim.1 ctr cos2 Dim.2 ctr cos2
## 1 | 3.250 | 1.012 0.019 0.097 | 0.942 0.026 0.084 |
## 2 | 2.008 | 0.805 0.012 0.161 | -0.047 0.000 0.001 |
## 3 | 3.315 | 0.852 0.014 0.066 | -0.663 0.013 0.040 |
## 4 | 1.841 | 1.040 0.020 0.319 | -0.978 0.028 0.282 |
## 5 | 2.029 | 1.579 0.047 0.606 | -0.136 0.001 0.004 |
## 6 | 3.075 | 1.616 0.049 0.276 | -0.091 0.000 0.001 |
## 7 | 2.559 | 1.031 0.020 0.162 | -1.282 0.048 0.251 |
## 8 | 2.441 | 0.762 0.011 0.097 | -0.048 0.000 0.000 |
## 9 | 3.198 | 0.304 0.002 0.009 | -1.388 0.056 0.188 |
## 10 | 4.496 | -2.847 0.153 0.401 | -1.205 0.042 0.072 |
## Dim.3 ctr cos2
## 1 2.302 0.202 0.502 |
## 2 -0.555 0.012 0.076 |
## 3 1.927 0.142 0.338 |
## 4 -0.786 0.024 0.182 |
## 5 0.188 0.001 0.009 |
## 6 -1.807 0.125 0.345 |
## 7 -0.480 0.009 0.035 |
## 8 -0.049 0.000 0.000 |
## 9 0.976 0.036 0.093 |
## 10 0.025 0.000 0.000 |
##
## Variables (the 10 first)
## Dim.1 ctr cos2 Dim.2 ctr cos2 Dim.3 ctr
## popularity | 0.002 0.000 0.000 | 0.095 0.596 0.009 | -0.617 33.386
## acousticness | -0.647 18.115 0.418 | 0.158 1.653 0.025 | -0.003 0.001
## danceability | 0.125 0.673 0.016 | 0.769 39.243 0.591 | -0.018 0.030
## duration_ms | -0.149 0.966 0.022 | -0.471 14.759 0.222 | 0.122 1.306
## energy | 0.872 32.917 0.760 | -0.267 4.720 0.071 | 0.015 0.021
## instrumentalness | -0.310 4.171 0.096 | -0.405 10.926 0.164 | 0.156 2.131
## liveness | 0.237 2.434 0.056 | -0.099 0.654 0.010 | 0.595 31.035
## loudness | 0.818 28.979 0.669 | -0.112 0.831 0.013 | -0.149 1.951
## speechiness | 0.083 0.298 0.007 | 0.482 15.415 0.232 | 0.578 29.288
## tempo | 0.159 1.093 0.025 | -0.254 4.297 0.065 | 0.084 0.612
## cos2
## popularity 0.381 |
## acousticness 0.000 |
## danceability 0.000 |
## duration_ms 0.015 |
## energy 0.000 |
## instrumentalness 0.024 |
## liveness 0.354 |
## loudness 0.022 |
## speechiness 0.334 |
## tempo 0.007 |
##
## Supplementary categories (the 10 first)
## Dist Dim.1 cos2 v.test Dim.2 cos2 v.test
## Dance | 0.678 | 0.273 0.162 3.998 | -0.339 0.249 -6.138
## HipHop | 0.843 | 0.024 0.001 0.382 | 0.592 0.492 11.692
## Pop | 0.735 | -0.082 0.013 -1.334 | 0.025 0.001 0.491
## Rap | 0.678 | 0.104 0.024 1.675 | 0.457 0.456 9.100
## Rock | 1.084 | -0.283 0.068 -4.512 | -0.786 0.526 -15.517
## *NSYNC | 3.488 | 2.905 0.694 1.912 | 0.196 0.003 0.159
## $uicideBoy$ | 2.143 | -0.348 0.026 -0.648 | 0.681 0.101 1.572
## 03 Greedo | 2.493 | 1.291 0.268 0.850 | 0.623 0.062 0.508
## 070 Shake | 2.524 | -1.786 0.501 -1.176 | 0.331 0.017 0.269
## 10cc | 5.564 | -3.790 0.464 -2.494 | -1.795 0.104 -1.463
## Dim.3 cos2 v.test
## Dance | 0.015 0.000 0.313 |
## HipHop | 0.500 0.351 11.333 |
## Pop | -0.534 0.528 -12.296 |
## Rap | 0.259 0.146 5.912 |
## Rock | -0.228 0.044 -5.167 |
## *NSYNC | 0.245 0.005 0.229 |
## $uicideBoy$ | 0.337 0.025 0.892 |
## 03 Greedo | -0.436 0.031 -0.408 |
## 070 Shake | 0.424 0.028 0.397 |
## 10cc | 0.606 0.012 0.567 |fviz_eig(tracks_pca, ncp = 11, addlabels = T, main = "Variance Explained by Dimensions")
non_numeric## genre artist_name key mode cluster
## 1 2 9 12 16fviz_pca_ind(tracks_pca, habillage = 1)
fviz_pca_var(tracks_pca) +
theme_minimal()
fviz_cluster(object = km_tracks, data = tracks_scale) +
theme_minimal()